52 by Bjørn T. Larsen1, Snorre Olaussen2, Bjørn Sundvoll3, and Michel Heeremans4 The Permo-Carboniferous through six stages and 65 million years

1 Det Norske Oljeselskp ASA, . E-mail: [email protected] 2 Eni Norge AS. E-mail: [email protected] 3 NHM, UiO. E-mail: [email protected] 4 Inst. for Geofag, UiO. E-mail: [email protected]

The Oslo Rift is the northernmost part of the Rotliegen- des basin system in Europe. The rift was formed by lithospheric stretching north of the Tornquist fault sys- tem and is related tectonically and in time to the last phase of the Variscan orogeny. The main graben form- ing period in the Oslo Region began in Late Carbonif- erous, culminating some 20–30 Ma later with extensive volcanism and rifting, and later with uplift and emplacement of major batholiths. It ended with a final termination of intrusions in the Early Triassic, some 65 Ma after the tectonic and magmatic onset. We divide the geological development of the rift into six stages. Sediments, even with marine incursions occur exclusively during the forerunner to rifting. The mag- matic products in the Oslo Rift vary in composition and are unevenly distributed through the six stages along the length of the structure.

Introduction

The Oslo Palaeorift (Figure 1) contributed to the onset of a pro- longed period of extensional faulting and volcanism in NW Europe, which lasted throughout the Late Palaeozoic and the Mesozoic eras. Widespread rifting and magmatism developed north of the foreland of the Variscan Orogen during the latest Carboniferous and contin- ued in some of the areas, like the Oslo Rift, all through the Permian period. We review the geological development of the Oslo Rift through its six stages of development (Ramberg and Larsen, 1978, Sundvoll et al., 1990, Olaussen et al., 1994), focusing on the four first—their lavas, sediments and tectonic structure, and briefly put it into the plate tectonic framework of NW Europe.

The Variscan orogeny, the Tornquist line and the Oslo Rift Figure 1 Simplified geological map of the Oslo Graben area. The Oslo Rift sediments exhibit great similarities to the Lower Brown—includes both volcanics, sediments and large dykes related Rotliegendes in the Northern European Permian Basin and in Katte- to the Oslo Graben; Carboniferous-Permian age. Red—large gat and may be regarded as a prolonged northern arm of the North- Permian batholithic intrusions. Small blue dots—Permian gabbroic ern Permian Basin. The Graben is the southern part of the intrusions. Green—Lower Palaeozoic sediments. Yellow—the Oslo Rift and is the link between the two tectonic systems (Heere- Caledonian thrust front. White—Pre-Cambrian basement rocks. mans et al., 2004). Abbreviations for different areas: Brum. = Brumunddal, Krok. = Recent reviews of post-Variscan tectonics in Western Europe Krokskogen, and for the caldera volcanoes; Øy = Øyangen, He = (McCann et al., 2006; Ziegler et al., 2006) have described the genetic Heggelia, Ni = , Bæ = Bærum, Gl = Glitrevann, Dr = relations and the timing between the Variscan orogeny and subse- , Sa = Sande, Hi = Hillestad and Ra = .

March 2008 53

Figure 2 Simplified tectonic overview of West Europe with the Variscan front, the Tornquist fault system and the Oslo Rift. Also shown are the pre-rift configurations with the Caledonian structures and the boundary of the Fennoscandian Craton. quent large, mostly NW-SE striking, wrench fault systems. The largest and northernmost is the Sorgenfrei-Tornquist Zone (Figure 2) that strikes across Scania (Skåne) into the (north of the Ringkøping-Fyn High), developing at least partly as a dextral strike- slip fault system. North of this fault, extensional stress fields devel- Figure 3 The graben segments and the graben polarity, the oped widespread rifting, being linked to the late stages of the master faults, the accommodation structures and the transfer fault orogeny and to the strike-slip faulting (Heeremans et al., 1997). in the Oslo Rift. Abbreviations of the structural nomenclature: formed both inside the orogen and in the foreland to the north, even R.F. = fault, S.H. = Solberg Horst, R.H.F = extending into the Fennoscandian Craton. The northernmost and the Randsfjorden-Hunnselv Fault, K.K.T.F. = Krokkleiva- largest of these structures was the Oslo Rift. Kjaglidalen Transfer Fault, E.T.F. = Ekeberg. Transfer Fault. Warr (2000) divided the development of the Variscan orogenic O.F. = Fault, and L.A.Z. = Langesund Accommodation system in NW Europe into four phases, separated both in time and in Zone. Li = , H = , D = Drammen, K = different areas. The last of the four phases is named the Asturian , M = Moss, S = , La = . phase and is generally Westphalian to Early Permian in age. Both Ziegler et al. (2006) and McCann et al. (2006) described it as the Finally, the offshore Skagerrak Graben represents the southern- consolidation phase of the Variscan Fold belt and gave an age most part of the Oslo Rift, and abuts towards the NW-SE trending 305Ma as the critical decline of the Variscan orogeny and the onset Sorgenfrei-Tornquist Zone in the south. The two and Vest- of rifting. Latest Carboniferous to earliest Permian was the time for fold graben segments form the classical Oslo Graben which is the onset of the Oslo Rift, leading up to its climax of both tectonic 220 km long and about 60 km wide. Adding the 100 km long Ren- and magmatic activity (Sundvoll et al., 1990; Heeremans et al., dalen Graben in the north, and the 180 km offshore Skagerrak 1997). Graben in the south makes the total length of Oslo Rift about 500 km. The Skagerrak Graben is broader than the other segments to the north, and is composed of several more or less overlapping The Oslo Rift architecture and grabens (Heeremans et al., 2004). The rift axis here strikes NE-SW, perpendicular to the Sorgenfrei-Tornquist fault system. nomenclature

The architecture of the Oslo Rift is very much like that of other well The petrogenesis of a high volcanicity rift known rift structures. Most have polarity off-set of grabens along the length of the rift axis, as described e.g. by Rosendahl (1987). The Larsen and Sundvoll (1984) summarized the Oslo Graben part of the Oslo Graben (Figure 3) was subdivided into two rift segments with Oslo Rift as a north-south trending Permo-Carboniferous high-vol- opposite subsidence polarity (Ramberg and Larsen, 1978). The canicity continental rift system, much like the recent East African Akershus Graben segment has an E-verging master fault (the Rands- rifts of Kenya and Ethiopia. The high volume of volcanics filling the fjord-Hunnselv Fault) to the north, while the Graben seg- rift is a feature common to both, and distinguishes them from other ment has a W-verging master fault (the Oslofjord Fault) to the south. continental low-volcanicity rifts such as the Baikal Rift and the These two half grabens have their accommodation zone around the Viking Graben. These two categories of rifts are useful descriptive city of Oslo, with a joining fault to the west of Oslo in the Kjagli- end-members (Barberi et al., 1982). dalen-Krokkleiva Transfer Fault (Heeremans et al., 1997). Today, A thorough analysis of the available data from the Oslo Rift we add the third Rendalen Graben segment to the north of the Aker- was undertaken by Neumann et al. (2004). They discussed the shus Graben, also with a west-verging master fault system, the Ren- magma origin and concluded that at least three mantle components dalen Fault (Skjeseth, 1963; Larsen et al., 2006). The accommoda- have contributed to the petrogenesis of the basaltic magmas, the old- tion, or transfer system, between the Akershus Graben and the Ren- est apparently being derived from an enriched mantle source. This dalen Graben is represented by the NE-SW trending Solberg Horst, source was most likely located in the lithospheric mantle and might beside lake Mjøsa. have been metasomatically altered by older carbonatitic fluid-rich

Episodes, Vol. 31, No. 1 54

melts (Anthony et al., 1989). The main mantle source for the basaltic magmatism was a prevalent depleted mantle. It may represent the composition of the base of the local lithospheric mantle, and the asthenosphere beneath, which partly melted in response to localized thinning of the lithosphere due to the extension. Anthony et al. (1989) also suggested an alternative scenario involving a mantle plume, with depleted characteris- tics, actively up-welling beneath the lithosphere. The most primitive lavas appear to involve low degree partial melting of one or more sublithospheric mantle sources. The rising man- tle-derived magmas were modified by shallow-level processes, including magmatic differentiation, general fractional crys- tallisation, magma mixing and lithospheric contamination that masked the geochemical signature of the mantle source. Large volumes of mantle-derived basaltic magma formed chambers near the Moho at c. 36 km depth. This also led to anatectic melting in the Precambrian host-rocks. Initial Sr isotopic ratios significantly above 0.7039 are typical of the syenitic and granitic rocks and imply influence of crustal contamination in the lower crust (Sundvoll et al., 1990). After 280 Ma, the rocks show a clear trend of increasing ini- tial ratios; mantle signature is only present in the larvikites and the rhomb-porphyry and basalt lavas. Sundvoll et al. (1990) interpreted the Sr-initial ratios to reflect the relative importance of mantle- versus crustal-derived melts. At c. 280 and 275 Ma, the magmatism became dominated by melts (syenitic and granitic) containing a larger crustal component. The mantle source had slowly become inactive, but mantle- derived magmas were still undergoing fractional crystallisa- tion in magma chambers in the lower crust giving rise to evolved rocks such as larvikites and late rhomb porphyry lavas, and to basaltic central volcanoes with shallower crustal magma chambers at c. 20 km depth. This termination of new mantle material into the Oslo Rift starts the mature stage of the rift. From this time, the slow tectonic and magmatic decay of the Oslo Rift started, and the rift segments devel- oped differently. This long “aftermath” period towards the final termination lasted about 15 million years in the Vestfold Graben, but as long as 35 million years in the Akershus Graben. The transition from the “mantle melting” to the “mature stage” and further to the “aftermath” is also reflected in a change in magma-tectonic style: (a) the magmatism migrated towards the central part of the graben, and (b) fis- Figure 4 An overview of the stratigraphy of the two first stages of the sure eruptions and normal faults gave way to central volca- development in the central part of the Oslo Graben. noes and, finally, to explosive volcanism and caldera- collapse. Mantle derived melts appear to have been present somewhat later under the Akershus Graben than under the Vestfold In the central part of the Oslo Graben, the up to 20 m thick Kol- Graben. This is in agreement with the proposed northwards propa- sås Formation unconformably overlies the folded Cambro-Silurian gation of the rift and its magmatism. sediments. The formation is dominated by red mudstones and sand- The asthenospheric mantle source of the most primitive mag- stones with subordinate greyish to greenish conglomerates, lime- mas bears the signature of being plume-related. The discussion on stones and rare anhydrite. The depositional environments have been the influence (if any) of a deep mantle plume on the magmatism in interpreted as floodplain, fluvial stream channel fill and shallow lake the Oslo Rift has been a controversy for about twenty years. Today, (Dons and Gyøry, 1967; Henningsmoen, 1978; Olaussen et al., opinions seem to be swinging in favour of a plume, apparently sup- 1994). Immature to mature calcrete soil profiles (paleosols) are ported by geophysical evidence (Torsvik et al., in press). recognised in the red overbank and floodplain facies (Olaussen, 1981). This, together with evaporite minerals suggest that arid and semi-arid conditions must have prevailed during the deposition of The six stages of development the Kolsås Formation. A 30 m thick similar unit occurs in the south- ernmost part of the Vestfold Graben (Olaussen and Dahlgren, 2007). The overlying 20 m thick Tanum Formation locally cuts down Stage 1: The proto-rift forerunner to rifting into the underlying Kolsås Formation. Also in the southern part of The Oslo Rift was initiated in the Late Carboniferous with the Oslo Graben, an unconformity separates a lower finer grained deposition of a thin carpet of sediments in the southern two-thirds of unit from an upper 60 m thick coarser section. The Tanum Formation what later was developed as the Oslo Graben. No record of sedimen- is well known for grey conglomerate beds, which reach up to tary rocks from this forerunner stage are observed NE of the city of 5 m in thickness and are often large-scale cross-stratified and Oslo, suggesting non-deposition. These sedimentary rocks are interbedded with medium to very coarse grained or pebbly sand- named the Group (Figure 4), which in the central part of the stones. Grey and finer grained sandstones, green and less common Graben is subdivided into three formations, of which the two first: red mudstone and limestone are other common lithologies. Dark the Kolsås Formation and the Tanum Formation belong to the proto- greyish plant-rich, minor coaly, mudstones have been observed. The rift stage. Tanum Formation has been interpreted as representing alluvial

March 2008 55

stream channels, floodplains and deltaic deposits (Dons and Gyøry, composition. These volcanic rocks are Si-undersaturated, mostly 1967; Olaussen et al., 1994). The coarser grained units are inter- basanites in the lower part and more alkaline basalts in the preted as braid plains and fan deltas deposits. Up to two meter thick upper half. Thin volcaniclastic sediments and pyroclastic products calcrete profiles with hardpan and some silcretes show evidence of are found between some of the lava flows. The Skien basalts, and development of mature soil profiles in an arid to semi-arid climate. A most likely also the basalts, came from an enriched man- thin marine limestone (sandy grainstone) overlies a fluvial or flu- tle source, potentially as a result of low degree partial melting within viomarine channel in the Kolsås area. This up to 2 m thick bed the garnet-bearing part of the mantle (Anthony et al., 1989). Thus, (Knabberud Limestone Member) is interpreted as a beach deposit the earliest and southermost basalts in the Oslo Rift indicate a low (Figure 4). Clasts of this lithology are also recognised in other cen- degree of partial melting in an asthenospheric mantle source, subse- tral parts of the Oslo Graben. Cross-stratified and horizontally lami- quently modified by lithospheric mantle and crustal components nated carbonate beds in Skien area resemble the same formation, (Neumann et al., 2004). A possible mantle plume or several smaller suggesting a widespread marine incursion in the Oslo Graben plumes have been inferred by Wilson et al. (2004). (Olaussen, 1981). The scattered occurrence of foraminifers The largest basalt province during Stage 2 is in the Vestfold/ (fusulinids), e.g., Novella evoluta mosquensis (Rauser) in the lime- Jeløya area, in the central Oslofjord area and continues all the way to stone suggests a late Bashkirian (Westaphalian A+B) to late northwest of Drammen. The thickness of the /Jeløya Moscovian (Westphalian D) age for the marine incursion. The fauna basalts varies from 1500 m at Jeløya, close to the Oslofjord Fault, to (freshwater mussels, fish and reptile remains) and flora in the upper- about 100 m at places in central northern Vestfold. The composition most part of the Tanum Formation and the overlying For- of this basalt suite is ordinary alkaline olivine basalts, and the thick- mation in the Asker area was suggested to be correlative with the ness of the individual flows varies from a few meters to c. 10 m. Lower Permian in (Holtedahl, 1931; Höeg, 1936). Some few flows of high-Ti basalts are found both at Jeløya and in It is suggested that this fossil-bearing unit in the Tanum Formation in Vestfold (Schou Jensen and Neumann 1985, Neumann et al., 1989). Asker overlies the Knabberud Limestone Member. Eager (1994) Many flows are aphyric, but flows with phenocrysts of olivine reviewed the freshwater mussels and suggested a correlation with (mostly altered), augitic clinopyroxene and are more Upper Pennsylvanian in North America. The flora has also been common. The basalts erupted from a series of composite fissure vol- reviewed and indicates a stage between Westphalian/Stephanian to canoes and from shield volcanoes, mostly as relatively thin com- Stephanian in northern Europe. The age of the Tanum Formation is pound lava flows, with nice examples of aa-lavas, pahoehoe-lavas also constrained by the radiometric age (291± 8Ma, Sundvoll et al. and columnar jointing. Volcaniclastic sediments are frequent (1990) in the central Oslo Graben) and 300±1 Ma, Corfu and Dahlgren between the flows; they are mostly thin and varing from fine sand (2007) in the southern Oslo Graben) of the first overlying lava. The (wind blown) to well-rounded coarse conglomerates. Pyroclastic formation was deposited prior to the major outpouring of lavas. products are also found, but are not frequent. The marine Knabbereud Limestone shows an affiliation with North of the basaltic lava field in Vestfold, no alkaline olivine carbonate platform units of the same age beneath the Barents Sea, basalts are found north of Drammen. However, in Lier and Asker, eastern and southern Europe. This implies that the Oslo area could we find sediments in this stratigraphic position. Alluvial debris flows have been flooded in the Late Carboniferous from both the east and were building out towards the north from the large basalt field in the south. The variable lithologies and facies variations observed in Vestfold. The sedimentary structures indicate a northward transport the Tanum Formation are consistent with deposition during a period and all clasts are of alkaline olivine basalts. These, up to 30 m thick, of major climatic and sea level changes, as recorded at the end of the red sediments are named the Skaugum Formation and occur only in Carboniferous, elsewhere. the central part of the Oslo rift. The unit is dominated by volcani- Spread over most of the Oslo Region, we find sill intrusions that clastics, varying from well sorted sandstones to coarse conglomer- are radiometrically dated to the Late Carboniferous (308–305 Ma), ates. Although there is an abrupt change in composition and in which indicate a “mid” Pennsylvanian age (Sundvoll et al., 1992; colour from grey to red between the Tanum and Skaugum forma- Sundvoll and Larsen, 1994). This corresponds to the “key-age” for tions, no major unconformity is observed. Also some of the freshwa- the start of magmatic activity, as also suggested by Ziegler et al. ter fauna and flora seen in the Tanum Formation continue into the (2006). The sill intrusions are primarily of syenitic composition Skaugum Formation. Inverse graded breccia occurs and is inter- (maenaite), but also basic camptonites occur. The thickness of the preted as deposition from debris flows. Together with similar conti- sills varies from cm-scale to more than 10 meters. nental depositional settings as in the Tanum Formation, this suggests a more proximal facies in the Skaugum Formation, probably repre- Stage 2: The initial rift and first basaltic volcanism senting an alluvial fan deposit (Olaussen et al., 1994). The Skaugum Stage two (Figure 4) exhibits only basaltic lava flows and an up Formation also exhibits calcrete soil profiles, suggesting a semi-arid to 20 m thick volcaniclastic unit, the Skaugum Formation of the to arid climate during deposition. Asker Group. The basalts from this stage show the whole suite of In Lier and Asker, on top of the Skaugum Formation volcani- compositions from highly silica undersaturated melilitites to over- clastic sediments, there occurs a single aphyric basaltic lava-flow. It saturated quartz tholeiites. The different basaltic lavas, however, are covers the whole of the central Oslo area, with a thickness varying found in different areas or provinces during this very early develop- from 5 to 20 m, and must have erupted as one simple flow from a fis- ment of the rifting. sure volcano. This flow, called the Kolsås basalt, has a quartz tholei- In the southernmost province of Brunlanes (Figure 1), a total itic composition and is the only basalt in the Oslo Rift of this com- stratigraphic thickness of 800 m of basaltic lavas is located. The position. North of this area, there is no indication of any eruption at Brunlanes basalt flows are relatively thin (~5–0.5 m), and all are sil- this stage in the rift. ica undersaturated melilitites and nephelinites. Recent U–Pb-dating has given the age 300 Ma for the earliest flows in the south (Corfu Stage 3: The rift climax, with rhomb porphyry and Dahlgren, 2007). We therefore assume that these highly fissure volcanoes Si-undersaturated basaltic lavas are the oldest in the rift system. These most strongly alkaline and undersaturated basalts erupted in When the change to Stage 3 started, a new era opened (Figure 4), an initial rift setting and are similar to the proto-basalts in the Kenya marked by the eruption of rhomb porphyry (RP) lavas. With this cli- rift (Williams, 1970 and 1971). In both rifts, they erupted closest to max stage of the Oslo Rift, the volume of eruptions increased. The the potential “hot spot” and very early during the rift evolution. eruptions of alkaline olivine basalts did not stop, but continued into To the northwest of Brunlanes, an equally thick suite of basaltic Stage 3. It continued and spread further to the north, but most likely lavas are exposed in the Skien- area called the Skien decreased in volume and intensity. The earliest rhomb porphyry basalts. The flows are slightly thicker than in Brunlanes and vary in lavas had extremely large volumes. Together they covered an area of

Episodes, Vol. 31, No. 1 56

the southern and central Oslo Graben from well north of Oslo city presumably terminated in an alluvial fan on the lowlands in the west. and to the outer part of the Oslofjord, i.e. about 10,000 km2. With an This, the Migartjern conglomerate, is uncomformably overlain by estimated thickness of at least 100 m, the total volume may have younger basalt and rhomb porphyry lavas (Larsen, 1978). In the sec- been as large as 1,000 km3. But most subsequent RP-flows were ond group, two major units are recognised: a minimum 1 km thick, much smaller than this first flow, varying in thickness from about 5 coarse rhomb porphyry conglomerate (Brøgger, 1900; Størmer, m to more than 100 m. Rhomb porphyry lavas are found all over the 1935; Larsen et al., 1978), banked against the eastern master fault in classical Oslo Graben, from Brumunddal in the north to Brunlanes in the Vestfold Graben, and a 800 m thick dune and wadi deposit in the the south, a distance of about 220 km. They are today concentrated southern part of the Rendalen Graben in Brumunddal. The first of in two large and a number of smaller lava plateau areas. The two these represents debris flows, stream and sheet floods in an alluvial large provinces are the Krokskogen lava plateau (about 400 km2) fan setting. The second, comprising the down-faulted red and yellow west of the city of Oslo and the Vestfold lava plateau (>1,000 km2) Brumund Formation, has both large scale, cross-stratified eolian in northern and central Vestfold. The stratigraphical thickness of the dune deposits and fluvial stream channel and flood plain deposits lavas in Vestfold is about 3 km, and about 75% are rhomb por- (Rosendahl, 1929). Recent mapping and sedimentological studies of phyries. At Krokskogen, the thickness is about 900 m with about the Brumund Formation have also recognised lacustrine limestones 80% rhomb porphyries. and calcrete paleosols (Lothe et al., 1999). These authors recorded The production rate of RP-flows in the different areas varied. In bedding to dip south-eastwards up to 60˚, and to decrease upwards, Vestfold, about 50 flows erupted during about 10 million years giv- and interpreted a syn-rift depositional environment. The sedimentary ing a production rate of about one flow every 250,000 years. At rocks which are preserved in the hanging-wall towards master faults Krokskogen, 20 flows erupted over 14 million years giving a pro- are a response to increasing topography in the graben. These sedi- duction rate of about one flow every 600,000 years (Sundvoll et al., mentary rock units are compared with the Rotliegendes deposits in 1990). The eruptions continued longer at Krokskogen than in Vest- the North Permian Basin. fold. And the first of the four flows in Brumunddal, in the southern The end of the climactic Stage 3 was marked by the emplace- part of the Rendalen Graben in the north, is much later than the first ment of major larvikitic batholiths. RP-flows further south. The rhomb porphyries have an intermediate composition with Stage 4: The mature rift, with central volcanoes and c. 55% SiO2 and with relatively high content of Na and K. phenocrysts constitute from c. 5 to 30% of the volume, and the caldera collapse feldspar crystals are mostly larger than 1 cm. The phenocryst The volcanic processes and the products again changed when feldspar is a ternary feldspar, zoned with a Ca-rich core and K-rich the rift-development approached Stage 4. Basaltic central volcanoes rim (Harnik, 1969). Such lavas, with a high content of Si and phe- nocrysts, would be expected to have a relatively high viscosity started to develop in many parts of the rift, which was now devel- (Sæther, 1962). But the volcanological “performance” of the RP- oped as a prominent structure with large faults bounding the grabens. lavas indicates that they flowed out quietly over large areas in Rhomb porphyries continued to erupt during this stage, but most “Hawaiian” style, accompanied by hardly any pyroclastic products. likely with a decreased intensity and volume, together with the new Consequently, the viscosity must have been relatively low. Two fac- formation of central volcanoes. tors likely explain this behaviour. Firstly, the temperature must have At tectonically strategic places all along the Oslo Graben, cen- been relatively high, being calculated to about 1050–1100˚C tral volcanoes started to form. Most were basaltic with slight varia- (Larsen, 1978), i.e. about the same temperature as that of an evolved tion in composition. They were mainly alkaline olivine basalts, but basalt with plagioclase phenocrysts. Secondly, there must have been also more Si-saturated transitional types occur as in the Glitrevann a high content of dissolved gases. Dissolved water and other gases volcano, north of Drammen. The most prominent tectonic setting of like halogens will decrease the viscosity. Recent analysis has shown some of these extrusions is a N-S trending string of volcanoes along that the content of Cl is low in the RP-lavas, but the content of F is the central axis of the Vestfold Graben, from Ramnes in the south to very high, 0.25 to 0.5%. Calculations using USGS software “Con- Glitrevann, NW of Drammen (Ramberg and Larsen, 1978). All cen- flow” (Mastin and Ghiorso, 2000) give viscosities of between 3 and tral volcanoes did not necessarily start to form at the same time and 5 Pa.s in the southern part in the southern part using temperatures it is possible that the southern ones started before those in the north. between 1050 and 1100˚C, gas-dissolution of 1–3% and phenocryst The diameter of the calderas at the present erosion level varies content between 3 and 10%. This is only slightly more viscous than between 12 km (e.g., at Ramnes and Nittedal) and 6 km (e.g., at “Kilauea basalt” from Hawaii and explains the rhomb porphyry Drammen). We infer from comparisons elsewhere that the original lavas volcanological behaviour. diameter of the pre-caldera volcanoes at their base was about three Outside the lava plateau, large, mostly N-S striking RP-dykes times the size, i.e. 36 to 18 km. Central volcanoes of similar setting occur over most of the area. Our impression is that the rare and huge with calderas are found in the Kenya Rift e.g. the Silali, RP lava flows erupted from large fissure dykes; the RP lavas are Suzwa and Menengai volcanoes (Williams, 1970; Baker et al., therefore classified as monogenetic fissure eruptions. Most of the 1971). By analogy with the East African Rift, the heights of the cen- RP-lavas have the appearance of simple flows; not compound flows tral volcanoes in the Oslo Rift are estimated to have been in the order like most of the basalts. By contrast, the basalts in Stage 2 probably of 1–1.5 km above the rift valley floor. Apparently, the Kenya and erupted from polygenetic fissure and central volcanoes as compound the Ethiopian rifts are today in a situation similar to that in the Oslo and thin flow systems. Both at Krokskogen and in Vestfold, we find Rift in the middle of Stage 4. The rift valley had formed, the central basalts interfingering with the RP-lavas. The basalts are mostly volcanoes were active and several of them had reached a mature alkali olivine basalts, but also undersaturated basanites occur. explosive caldera-forming stage. Apart from the extrusions, some sedimentary units are pre- All the large central volcanoes inside the Oslo Rift seemed to served as remnants of the original basin fill within the Oslo Graben have matured petrologically. Olivine, augitic clinopyroxene and pla- (Olaussen et al., 1994). They can be grouped into two types: sedi- gioclase crystallized out of the basaltic magma leaving a felsic, mentary rocks preserved between the lava flows, and thick (up to mildly alkaline magma as the residual melt product. The latter was 1 km) units above the lava flows. The first group include fluvial and mostly trachytic in composition, but some also developed to alkaline alluvial fan deposits and thin calcrete paleosols. Units with finer rhyolites. The magma-products finally erupted explosively during grained ripple-laminated sandstones and mudstones with stroma- the caldera formation. At the present erosion level in the Oslo tolithic limestone provide evidence for the development of ponds or Graben, felsic ring-dykes and central domes are exposed together smaller lakes. Up to 1 m thick, well sorted and rounded, large scale with large and small remnants of ignimbritic effusives mixed cross-stratified fine grained sandstones are interpreted as eolian together with basalts and other pre-caldera products. Only in Vest- dunes. A 400 m thick canyon-fill in the Krokskogen lava plateau, fold do we find thick and widespread ignimbrites in the upper part of

March 2008 57

the rhomb porphyry succession outside the calderas, most likely References erupted from the nearest calderas. The age of these trachytic ignimbrites (T1 and T2) in inner Vestfold has been dated to 288 Ma Anthony, E.Y., Segalstad, T.V., and Neumann, E.-R., 1989, An unusual man- and 285 Ma, respectively (Sundvoll and Larsen 1990). Even though tle source region for nephelinites from the Oslo Rift, Norway: Geochim. these datings have limited precision, they indicate that the explosive et Cosmochim. Acta, 53, 1067–1076. eruptions in Vestfold are older than the ones further north, and that Baker, B.H., Williams, L.A.J., Miller, J.A., and Fitch, F.J., 1971, Sequence the different development stages distinguished in the Oslo Graben, and geochronology of the Kenya rift volcanics: Tectonophysics, 11, 191–215. were not synchronous, but started in the south. At Krokskogen, we Barberi, F., Santacroce, R., and Varet, J., 1982. Chemical aspects of rift mag- lack significant dating among the caldera-related extrusive products, matism, in G. Palmason ed., Continental and oceanic rifts: American but the intrusions (ring-dykes and central intrusions) in the caldera Geophysical Union, 223–258. cluster formed at c. 270 Ma, implying that the explosions from these Brøgger, W.C., 1900, Konglomerater i Kristianiafeltet. I. Om porfyrkon- calderas (e.g., the Øyangen caldera) most likely were significantly glomeratet paa ørekken Revlingen-Søstrene, en sedimentær formation fra younger than in Vestfold. Since ignimbrites have not been observed Kristaniafeltet: Nyt Mag. Naturv, 38, 29–64. in any part of the Krokskogen plateau, and the uppermost RP-lava Corfu, F. and Dahlgren, S., 2007, Pervoskite U-Pb ages and the Pb isotopic was dated to c. 276 Ma, we have an interval of about 10 Ma between composition of alkaline volcanism initiating the Permo-Carboniferous Oslo Rift: Earth Planet. Sci. Lett. 265, 256–269. the first Vestfold caldera explosions and those at Krokskogen. Dons, J.A., and Györy, E., 1967, Permian sediments, lavas and faults in the The youngest sedimentary units preserved inside the Oslo Rift Kolsås area W of Oslo: Norsk Geol. Tidsskr. 47, 57–77. calderas are red breccias, sandstones and thin laminated mudstones, Eager, R.M.C., 1994, Non-marine bivalve assemblage in the Asker Group, often associated with pyroclastic rocks. These units are interpreted to Oslo Graben and its correlation with late Pennsylvanian assemblages be of alluvial and lacustrine origin, and the best preserved are inside from North America: Jour. Geol Soc., , 252, 669–680. the Nittedal caldera (Naterstad, 1978). Harnik, A.B., 1969, Strukturelle Zustände in den Anothoklasen der Small alkaline gabbros were intruded during Stage 4 at c. 265 Rhombenporphyre des Oslogebietes: Schweizerische Min. Petr. Mitt. 49, 509–567. Ma and are located along N-S tectonic lineaments like at , Heeremans, M., Larsen, B.T., and Stel, H., 1997, Paleostress reconstruction and in clusters as in the central Oslofjord area (Neumann et al., from kinematic indicators in the Oslo Graben, : new 1985). They are only from 1 km to 100 m in diameter, sometimes constraints on the mode of rifting: Tectonophysics, 266, 55–79. exhibit layering and represent basaltic magma chambers that existed Heeremans, M., Kirstein, L., Larsen, B.T., and Timmerman, M., 2000, The below smaller basalt volcanoes at the same time as the larger ones structural evolution of Permo-Carboniferous dykes and sills in NW that developed to caldera volcanoes (Steinlein, 1981). Europe: a multidisciplinary approach: Geoscience 2000 Abstracts, p66. The large batholithic Drammen and a slightly smaller Heeremans, M., Faleide, J.I., and Larsen B.T., 2004, Late Carboniferous-Per- mian of NW Europe: an introduction to a new regional map, in Permo- body further to the north, the Finnemarka granite, intruded into the Carboniferous magmatism and Rifting in Europe: Geol. Soc. London, northern part of the Vestfold Graben, slightly south of the transfer Spec. Publ. 223, 75–88. system that separates the latter from the Akershus Graben. Henningsmoen, G., 1978, Sedimentary rocks associated with the lava series, in J.A. Dons and B.T. Larsen eds, The Oslo Paleorift: Norges Geol. Stage 5: The magmatic aftermath, with major Unders. 337, 17–24. Holtedahl, O., 1931, Jungpaläozoische Fossilen im Oslogebiete: Norsk Geol. syenitic batholiths. Tidsskr. 12, 323–340. Höeg, O.A., 1936. The Lower Permian flora of the Oslo region: Norsk Geol. After the youngest intrusions related to the development of the Tidsskr. 16, 1–43. Stage 4 calderas (c. 266 Ma, Sundvoll and Larsen, 1990), a set of Larsen, B.T., 1978. Krokskogen lava area, in J.A. Dons and B.T. Larsen large batholiths developed mostly in two areas—in inner Vestfold (Eds.) A review and guide to excursions: Norges Geol. Unders., Bull. (west of the lava areas and north of the older larvikite batholiths) and 337, 143–162. in the and areas, north of Oslo (Figure 1). Proba- Larsen, B.T., Ramberg, I.B. and Schou Jensen, E., 1978. Excursion 3 Cen- bly due to the present erosional level, no extrusions or sediments can tral part of the Osloford, in J.A. Dons and B.T. Larsen eds, A review and be linked to this Stage 5 of igneous activity, which lasted from c. 265 guide to excursions: Norges Geol. Unders., Bull. 337, 105–124. Larsen, B.T. and Sundvoll, B., 1984. The Oslo Graben: A Passive High Vol- to 255 Ma. These batholiths are mostly alkali syenitic to alkali canicity Continental Rift: EOS 65, no 45, 1084. granitic in composition and are called Nordmarkites and Ekerites. Larsen, B.T., Olaussen, S., Sundvoll, B. and Heeremans, M., 2006. (English Several are intruded into the lava plateau, calderas and earlier edition 2008). Vulkaner, forkastninger og ørkenklima, in I.B. Ramberg, I. batholiths, and partly destroyed these structures. Bryhni og A. Nøttvedt eds, Landet blir til: Norsk Geologisk Forening. p. 284–327. Lothe, A. E., Haugen, M., Gabrielsen, R., Larsen, B.T., Olaussen, S. and Tal- Stage 6: Rift termination, with the youngest small bot, M., 1999. Structural and sedimentological aspects of the Early Per- granite intrusions mian. Brumunddal Sandstone, northern Oslo Graben (abstract, Geol. soc, Norway): Geonytt. p. 68. This last stage of magmatic activity occurred in two separate Mastin, L.G. and Ghiorso, M.S., 2000, A Numerical Program for Steady- areas, both north of the city of Oslo: in the Tryvann area in the north- State Flow of Magma-Gas Mixtures through vertical erupive conduits: ern hills of Oslo, and further to the north in Hurdal. The intrusions USGS Open-File Report 00-209. 61 pp (http://vulcan.wr.usgs.gov/Pro- are granitic in composition with ages from 250 to 245 Ma (Sundvoll jects/ Mastin/Publications/OFR00-209/conflow.htm). et al., 1990). Younger dikes also exist, primarily in the northern Oslo McCann, T., Pascal, C., Timmermann, M.J., Krzywiec, P., López-Gómez, J., Wetzel, A., Krawczyk, C.M., Rieke, H. and LaMarche, J., 2006, Post- Graben (Torsvik et al., 1998, Heeremans et al., 2000). Variscan (end Carboniferous-Early Permian) basin evolution in Western abd Central Europe, in D.G. Gee and R.A. Stephensson eds, European Epilogue Lithosphere Dynamics: Geol. Soc. London, Memoir 32, 355–388. Naterstad, J., 1978, The Nittedal Cauldron (Alnsjøen Area), in J.A. Dons and The geological development of the high-volcanicity continental B.T. Larsen eds, A review and guide to excursions: Norges Geol. Unders. Oslo Rift is described through six stages in Late Carboniferous and Bull. 337, 99–103. all through the Permian. Such palaeorifts structures are rarely Neumann, E.-R., Larsen, B.T. and Sundvoll, B., 1985, Compositional varia- exposed due to post-rift cooling, subsidence and younger sedimenta- tions among gabbroic intrusions in the Oslo Rift: Lithos, 18, 35–59. Neumann, E.-R., Wilson, M., Heeremans, M., Spencer, E.A., Obst, K., Tim- tion. Though the Oslo Rift has been studied for nearly 200 years and merman, M.J. and Kirstein, L., 2004, Carboniferous-Permian rifting and has the easiest possible access, with the capital, Oslo, located in the magmatuism in southern , the North Sea and northern Ger- middle of the structure, many questions remain unresolved, awaiting many: a review, in Permo-Carboniferous magmatism and Rifting in new investigations. Europe: Geol. Soc. London, Spec. Publ. 223, 11–40.

Episodes, Vol. 31, No. 1 58

Olaussen, S., 1981, Marine incursion in Upper Paleozoic sedimentary rocks Bjørn T. Larsen is educated from the of the Oslo Region, Southern Norway: Geol. Mag., 118, 381–388. Universities of and Oslo, with Olaussen, S., Larsen, B.T. and Steel, R., 1994, The Upper Carboniferous-Per- a cand. real. graduation in Oslo in mian Oslo rift; Basin fill in relation to tectonic development, in Embry, 1979. He has worked at the University A.F., Beauchamp, B. and Glass, D.J. eds, Pangea, Global environments of Oslo with the Oslo Graben studies and resources: Can. Soc. Petrol Geol. Memoir 17, 175–198. Olaussen, S. and Dahlgren, S., 2007, Environmental setting of the southern- until he in 1986 started working for most outcrop of the Carboniferous in the Oslo Rift. An arid syn-tectonic the Norwegian oil companies. He lacustrine and alluvial deposit with possible marine incursions, in worked for Statoil, Hydro and Saga Nakrem, H.A.(ed), Vinterkonferansen 2007: Abtracts and Proceedings of with Norwegian and international the Geological Society of Norway. No 1, 2007, 70–71. exploration until 2000 when he started Ramberg, I.B. and Larsen, B.T., 1978, Tectonomagmatic evolution, in J.A. as a consultant. He has continued Dons and B.T. Larsen, eds, The Oslo Paleorift: Norges Geol. Unders. working in the Oslo Graben together 337, 55–73. with co-workers from University of Ramberg, I.B. and Spjeldnæs. N., 1978, The tectonic history of the Oslo Oslo and the Geological Survey of Region, in E-R. Neumann and Ramberg, eds, Tectonics and Geophysics of Continental Rifts: Reidel Dordrecht, 167–194. Norway (NGU) parallel with his oil- Rosendahl, B.R., 1987, Architecture of continental rifts with special refer- industry affiliations, and has arranged ence to East Africa: Annual Review, Earth and Planetary Science Letters, several professional field trips in the 5, 445–503. Oslo Graben. Rosendahl, H., 1929, The porphyry-sandstone sequence in Brummundal: Norsk Geol. Tidsskr. 10, 367–438. Schou Jensen, E. and Neumann, E.-R., 1988, Volcanic rocks at Jeløya, cen- Snorre Olaussen is educated from tral Oslo Region: the mafic lavas: Norsk Geol. Tidsskr. 68, 289–308. the Universities of Oslo and Bergen Skjeseth, S., 1963, Contribution to of the Mjøsa District and the clas- and graduated with his MSc in 1979 sical Sparagmite area in southern Norway: Norges Geol. Unders. 220. and his Ph.D. in 1984. He joined the 226 pp. Norwegian oil industry in 1980, and Steinlein, O., 1981, En petrologisk og geokjemisk undersøkelse av lagdelte, has worked for Statoil, Saga and gabbroide vulkanplugger i , Oslo-graben. Univ. i Oslo, cand. Real. Hydro; as a chief geologist in Saga thesis, 93 pp. Størmer, L., 1935, Contribution to the geology of the southern part of the and Hydro. Since 2000 he has been Oslofjord: The rhomb porphyry conglomerate with remarks on younger working as a new venture/regional tectonics: Norsk Geol. Tidskr. 15. 43–133. team leader for Eni Norge. He has Sundvoll, B., Neumann, E.-R., Larsen, B.T. and Tuen, E., 1990, Age rela- continued his arrangements in the tions among Oslo Rift magmatic rocks: implications for tectonic and Oslo Region and has arranged magmatic modelling. Tectonophysics 178, 67–87. several professional field-trips. Sundvoll, B., Larsen, B.T. and Wandås, B., 1992, Early magmatic phase in the Oslo Rift and its related stress regime, Tectonophysics, 208, 37–54. Sundvoll, B. and Larsen, B.T., 1993, Rb-Sr and Sm-Nd relationship in dyke and sill intrusions in the Oslo Rift and related areas: Norges Geol. Unders. Bull. 425, 23–40. Bjørn Sundvoll is educated from the Sundvoll, B. and Larsen, B.T., 1994, Architecture and early evolution of the University of Oslo and received his Oslo Rift: Tectonophysics. 240, 173–189. cand.real. thesis in isotope chemistry. Sæther, E., 1962, Studies of complex of the Oslo Region, Since 1974, he worked for the XVIII, General investigation of the igneous rocks in the area north of Geological Survey of Norway with Oslo. Skrifter Norske Videnskaps Akademi I Oslo, Matematisk- the working address at the naturvitenskapelig klasse: Ny serie. No. 1, 184pp. Geological Museum at University of Torsvik, T.H., Eide, E.A., Meert J.G., Smethurst, M.A., and Walderhaug, 40 39 Oslo. He was part of the team that H.J., 1998, The Oslo Rift: New Palaeomagnetic an Ar/ Ar age con- established the isotope laboratory in straints: J. Geophys. Int. 135, 1045–1059. Torsvik, T.H., Smethurst, M.A., Bruke, K. and Steinberger, B., 2007, Long Oslo. In 1995, he defended his Dr. term stability in Deep Mantle structure: Evidence from the ~300 Ma Philos. thesis on isotope chemistry Skagerrak-Centred Large Igneous Province (the SCLIP): Earth Planet. and radiometric datings of Oslo Sci. Lett. (submitted). Graben rocks. Warr, L.N., 2000, The Variscan Orogeny: The welding of Pangea, in Wood- cock, N. and Strachan, R. eds, Geological History of Britain and Ireland: Blackwell, Oxford, 271–391. Williams, L.A.J., 1970, The volcanics of the Gregory Rift Valley, East Africa: Bull. Volcanologique, 3, 439–465. Michel Heeremans is a university Williams, L.A.J., 1971, The Kenya Rift volcanics: A note on volume and lecturer in petroleum geology at the chemical composition: Tectonophysics, 15, 83–96. University of Oslo. His interest in the Wilson, M., Neumann, E.-R., Davis, G.R., Timmerman, M.J. Heeremans, Oslo Graben started during his PhD M., and Larsen B.T., 2004, Permo-Carboniferous magmatism and rifting at the Vrije University, . in Europe: introduction, in Permo-Carboniferous magmatism and Rifting After his graduation in 1997, he in Europe: Geol. Soc. London, Spec. Publ. 223. 1–10. moved to Norway where he con- Ziegler, P.A., Schumacher, M.E., Dèzes, P., Van Wees, J.-D., and Cloething, tinued his work, but now focusing on S., 2006, Post-Variscan evolution of the lithosphere in the area of the European Cenozoic Rift System, in Gee, D.G. and Stephenson, R.A. eds, the Carboniferous-Permian develop- European Lithosphere Dynamics: Geol. Soc. London, Memoir 32, ment of NW Europe. His current 97–112. work focuses mostly on education and database support.

March 2008